Comments

• Michael

  2012-12-01 18:46:54

  This is a *massive* opportunity for open source advocates! Let the OEMs know you want a Linux-friendly box. Let them know you install Linux and do not use Windows.
  
  How will they react if enough people do this? It is obvious: they will ship Linux-friendly boxes or, better yet, ship Linux pre-installed. The challenge there is for them to decide what distro and to be able to do support, etc. Would it be worth it... only if we open source advocates let the OEMs know.
  
  So call Dell. Call HP. Let them know what you want. If there really is significant demand for Linux on the desktop next year really could be the year of desktop Linux.

• mjg59

  2012-12-01 20:12:13

  That's one article that incorrectly claims "If you want to run Linux or Unix or anything that wasn't sold by Microsoft, you need to disable the Secure Boot feature.", one article that's talking about a machine that has UEFI but no legacy BIOS compatibility (and so will only boot modern Linux distributions that have UEFI support) and one that's actually about secure boot but is mostly just saying that Fedora has better UEFI support than most other distributions.

• Dr. Roy Schestowitz

  2012-12-01 22:38:04

  You are dismissing real criticisms in your comment (also in Twitter), but this dismissal does no justice to those who genuinely complain.
  
  First of all, for the first case, one needs to have BIOS skills (which few people out there have, percentage-wise wrt the general population), one cannot use older distributions (does that sound OK to you? Is this freedom?), and one is better off using 'supported' distros like Fedora, which I guess works well for you because you're a Fedora developer.
  
  What about the small players? Well, as you point out, they could adapt, but why should they adapt by chasing Microsoft's demands (from OEMs)? For Microsoft to do what it does clearly takes some abuse of monopoly power. This bootloader you present reduces choice, still. Who needs UEFI anyway? The only security it offers is Microsoft's financial security. Even Torvalds mocked its claims of added security. So how can Microsoft justify this to regulators? It can't. Here is what the Microsoft-funded ($100m in 2010) SUSE is doing: "Suse's bootloader design involves the bootloader having its own key database, distinct from those provided by the UEFI specification. The bootloader will execute any second stage bootloaders signed with a key in that database. Since the bootloader is in charge of its own key enrolment, the bootloader is free to impose its own policy - including enrolling new keys off a filesystem."
  
  UEFI hardly improves Linux in any way; it's just a competition disruptor and based on users' accounts, it's very effective at that.
  
  Larabel says:
  
  

  Matthew Garrett has shared that he's finally published his shim boot-loader for dealing with UEFI SecureBoot that makes it easier for the smaller Linux distributions to deal with this "secure" technology. Using this shim boot-loader is already signed with a Microsoft key so the smaller Linux distributions and other independent parties don't have to worry about obtaining a key from Microsoft.

  
  
  You do realise that getting keys from Microsoft defeats much of the purpose of Free software, right? You do know that Germany's government (I'm a German citizen, mind you) bans UEFI-infected PCs, right? And why? Because, hypothetically, if a war erupts some day in the future, machines can be disabled at software and even hardware level, thanks to Microsoft. The latter can affect BSD and Linux too.
  
  Software freedom, which I assume you don't love so much because I see you bash the FSF, is not some 'hippie' hype du jour, it goes a lot deeper than this and control is required for real security. UEFI is only reducing security. It gives another vendor (or vendors) the ability to abduct one's machine.

• Michael

  2012-12-01 22:49:19

  The OEMs and MS owe *nothing* to the desktop Linux ecosystem.
  
  If their really isa large demand for desktop Linux then this is a *perfect* opportunity to let the OEMs know. No more can they use the excuse that you can just buy the system and install what you want.
  
  If there is demand they will meet it. You can count on that. This is a perfect chance to show that desktop Linux has earned its place in the desktop market; that it should be pre-installed.
  
  That is if it *had* earned such a place. Sadly that seems to not be the case. So you whine about your scary boogieman. Stop. Work to make desktop Linux earn the spot you want it to have.

• mjg59

  2012-12-02 01:12:25

  "First of all, for the first case, one needs to have BIOS skills"
  
  Incorrect.
  
  "one cannot use older distributions"
  
  Most of which won't support the hardware properly anyway, but yes, I'm unhappy about that.
  
  "one is better off using ‘supported’ distros like Fedora, which I guess works well for you because you’re a Fedora developer."
  
  I've no idea what a supported distribution is - if you mean that hardware compatibility is improved by running distributions that have significant developer effort, sure.
  
  "For Microsoft to do what it does clearly takes some abuse of monopoly power."
  
  I'd love to think so, but all the lawyers I've spoken to disagree.
  
  "The only security it offers is Microsoft’s financial security."
  
  Incorrect.
  
  "Here is what the Microsoft-funded ($100m in 2010) SUSE is doing"
  
  Code derived from my code, which has then been merged back in. Behaving like good free software developers, in other words.
  
  "UEFI hardly improves Linux in any way"
  
  It improves it in numerous ways.
  
  "You do realise that getting keys from Microsoft defeats much of the purpose of Free software, right?"
  
  Sure, hence why I've put all this work into helping develop a solution that ensures the end-user still has the freedom to boot their own bootloader and kernel.
  
  "You do know that Germany’s government (I’m a German citizen, mind you) bans UEFI-infected PCs, right?"
  
  No it doesn't. A German government white paper says that Secure Boot must be disabled by default (a position I agree with - see the Red Hat/Ubuntu white paper on this from over a year ago), but doesn't require that systems have any legacy BIOS support. UEFI-only machines can meet those requirements.
  
  "Because, hypothetically, if a war erupts some day in the future, machines can be disabled at software and even hardware level, thanks to Microsoft."
  
  Incorrect.
  
  "Software freedom, which I assume you don’t love so much because I see you bash the FSF"
  
  I was a speaker at Libreplanet earlier this year. I've worked closely with the FSF to ensure that proposed solutions were GPL compatible. I was drinking with some FSF staff members a couple of weeks ago. I've worked with various vendors to ensure that they provide source code to their customers. I'm a firm believer in software freedom and I think the FSF do great work.
  
  "UEFI is only reducing security. It gives another vendor (or vendors) the ability to abduct one’s machine."
  
  Incorrect.
  
  You don't appear to understand any of the technical or implementation details of this. You should really do some research before making obviously incorrect assertions.

• Michael

  2012-12-02 01:25:44

  Thank you for words of sanity on this issue and on this site.
  
  Roy is in many ways a good man. He truly believes in what he does. He allows me, a harsh critic, to respond to him over and over - as far as I know he has never removed a single comment of mine even though I rarely agree with him and often strongly disagree.
  
  With that said, he often knows little of what he is writing about, thinks Microsoft and Apple are pure evil (but Google and Samsung, apparently, are the good guys), and is very paranoid about many tech topics.

• Dr. Roy Schestowitz

  2012-12-02 01:33:59

  Some of these answers, starting with the first one, I disagree with. For instance, in many cases people will need to access the BIOS. They already do. I gave examples.
  
  Jumping to the last (which I remember best as it's fresh in my mind): Can you offer a URL that's a refutation of UEFI as an obsolescence tool? When UEFI cannot be disabled (e.g. various ARM boards) this is a real problem. Seeing the whole UEFI/ARM controversy helps here (Microsoft's requirements). The silicon/equivalent is metaphorically speaking hostile towards some software. Before you argue that it's no threat to national security, see:
  
  http://en.wikipedia.org/wiki/Stuxnet
  
  "Stuxnet is a highly sophisticated computer worm. Discovered in June 2010, Stuxnet initially spreads via Microsoft Windows, and targets Siemens industrial software and equipment."
  
  This was a tool of conquest. It took down operations strategically, by governments. Anything that lets the BIOS process be updated remotely is equally risky.
  
  I retract my statement that you dislike the FSF, but I did see you making a derogatory statement in relation to them once.
  
  I am not looking for trouble or any unnecessary provocations. I am truly concerned that UEFI takes everyone -- Windows users too -- a step back (like the Internet being further controlled by authorities and corporations, at the expense of peers), let alone the anti-competitive aspects.

• mjg59

  2012-12-02 01:40:48

  They need to access the firmware because the distribution they're using hasn't implemented support yet. Installing Ubuntu 12.10, for instance, requires no firmware interaction.
  
  I'm not sure what you mean by it being impossible to disable UEFI. You can't disable the BIOS on BIOS-only systems, and you can't disable UEFI on UEFI-only systems. How is that a problem? It's been possible to perform BIOS updates remotely for a long time now - the majority of platforms support a runtime mechanism for doing so.

• Michael

  2012-12-02 01:46:38

  Roy, you seem to think anyone who is not jumping up and down to support your BS is somehow against open source software... I know you made that accusation against me, even though I am very much for it and very appreciative of the work the open source community does. I use open source software daily.

• Dr. Roy Schestowitz

  2012-12-02 02:10:03

  If we have fewer distributions (and/or version) supported, how does that make UEFI benign or beneficial? Name an advantage for it.
  
  On UEFI-only systems you could use malicious remote updates to altogether render a machine "bricked".

• Jose_X

  2012-12-02 03:54:34

  mjg59, >> “For Microsoft to do what it does clearly takes some abuse of monopoly power.”
  
  >> I’d love to think so, but all the lawyers I’ve spoken to disagree.
  
  I was reading on uefi basics from wikipedia.
  
  Can you elaborate on what you were told is not an antitrust violation wrt Microsoft using their OS/app platform leverage to force ARM vendors to require MS signed key access if they want the Win OS? This seems like a case of large market power in OS/app being used to decide hardware requirements that would be detrimental to all competing OS, in particular, low cost OS and custom OS like Linux? The weight of one player is used to put all other competitors in a disadvantageous position and in a way that hurts consumers (significantly fewer options, in particular, lower cost options).
  
  I don't know enough about the $99 key, but unless we can buy one and share it with everyone through open source, that would be a major impediment.
  
  Note that the Wikipedia article says that after criticisms against Microsoft, they stated that on intel arch they would not require key signing (a position you agree is important) but they do require it on ARM, creating the antitrust issue I would think (ie, for ARM platforms).
  
  Also, can you comment on the following from http://www.zdnet.com/blog/open-source/linus-torvalds-on-windows-8-uefi-and-fedora/11187 > Fedora explored other options. "An alternative was producing some sort of overall Linux key. It turns out that this is also difficult, since it would mean finding an entity who was willing to take responsibility for managing signing or key distribution. That means having the ability to keep the root key absolutely secure and perform adequate validation of people asking for signing.
  
  Why would you need to keep anything secure? Can't you have a generic key in order to undo the layer of boot security basically as if we were back to BIOS only, which is how Linux has been used successfully by many people for many years?
  
  I don't know what UEFI does precisely. Any details that would help clarify these questions I have would be appreciated.

• Dr. Roy Schestowitz

  2012-12-03 18:22:43

  From http://mjg59.dreamwidth.org/20303.html
  
  

  Date: 2012-12-01 07:33 am (UTC) From: [personal profile] mjg59 Sure. I don't believe that it's possible without Windows for the final upload - it really needs Silverlight. Running IE under Wine may be sufficient, but I couldn't be bothered. Anyway. 1. Go to sysdev.microsoft.com and log in with a Live account. 2. Follow the link to the Verisign (now Symantec) page for creating a new company account. Ignore the use of the word company - you can do this as an individual. 3. Follow the instructions and purchase an individual key for code signing. You'll be emailed a form to attach a copy of your notarised ID to, so get that filled in and signed and send them back a copy by email. 4. Export the key from your browser as a .p12 file. 5. Go back to sysdev.microsoft.com and download the zip file containing winsign.exe. Use pesign or sbsign and the key you exported to sign this file, and then upload it to sysdev.microsoft.com to enable your account. 6. Sign the legal agreements - this just involves you typing your name into a box. 7. Put the file you want to get signed into a cab file. lcab will do this, 8. Sign the cab file with your Verisign key. osslsigncode will do this. 9. Upload the file to sysdev.microsoft.com. The uploader is Silverlight for no obviously good reason. 10. Wait for the upload to be processed. I think this happens a couple of times a week, so be prepared to wait a few days (I had to) 11. You'll get an email when signing is complete. Download the cab file and use cabextract to retrieve your signed binary. Total cost is $99 plus however much it costs to get something notarised where you are

• Michael

  2012-12-03 20:09:13

  You hate the DOJ because they were stopping people from breaking the law by downloading illegal content.
  
  Amazing.

• mjg59

  2012-12-03 22:21:29

  Jose,
  
  A huge number of ARM vendors ship systems that are locked down to running specific operating systems. Most Android devices have locked bootloaders. All Apple ones do. I think this is an incredibly anti-user design trend, and I applaud everyone who's put work into breaking those bootloader locks, but the truth is that Microsoft are behaving in exactly the same way as many other ARM vendors and nobody's suing any of them.
  
  Roy,
  
  I agree that it's a shame that some existing distributions won't run on Secure Boot systems, but old versions of Debian are completely useless on my current Thinkpad even without UEFI. The industry is based on a constant stream of new and incompatible devices, and if you want to run old distributions on that new hardware you have to jump through new hoops.

• mcinsand

  2012-12-03 18:06:34

  One of the duopoly's greatest fears is that the Antitrust authorities might actually waken from what appears to be a very, very deep coma. After all, as ruled by the 9th courm of appeals and upheld by the US Supreme Court in DG versus Digidyne: "the Ninth Circuit Court of Appeals has held a computer systems manufacturer's method of marketing its hardware and software to be a per se illegal tying arrangement in violation of both the Sherman and Clayton Antitrust Acts." The method of marketing was that the RDOS operating system was only to be sold with DG hardware (Nova and Eclipse minicomputers, in particular). One half of the Applesoft duopoly has flouted this ruling for decades with impunity and, now, the other half is looking for the same. Linking hardware to OS is just as ridiculous as telling Ford car drivers that they can only use Ford fuel, oil, and consumables, and it is just as illegal.
  
  I also like this quote from 734 F.2d 1336: "the Supreme Court again made it clear that a tying arrangement is illegal per se if the seller of the tying product has the capacity to force some buyers to purchase a tied product they do not want or would have preferred to purchase elsewhere." In other words, based on the ruling, if people wanted to purchase the Surface hardware without the toy OS, then forcing them to purchase Windows 8 is illegal. If someone wants to buy their own hardware and a crippled version of BSD, then Apple should be letting them purchase and use a copy of OSX.
  
  Maybe the word 'coma' that I used is overly optimistic, on second thoughts. Perhaps an RIP wreath to the US DOJ would be more appropriate.
  
  I think another problem driving UEFI is that installing Windows is not what it used to be. Hardware support has shifted drastically, and Windows' support is truly horrible. In the past few years, I have had far less trouble with even wireless drivers than with Windows. In fact, it was a wireless driver failure that finally motivated one friend to try Ubuntu last year. The computer worked flawlessly until yesterday, when she decided to boot into Windows 7. The Microsoft half of Applesoft might still have better hardware support than the Apple half (if you can't install an OS on a whitebox purchased from Tigerdirect, your local shop, Aldi, or whever, then the OS has no hardware support). However, Windows is rapidly sinking to OSX' level.
  
  Although security is just a weak excuse, UEFI is yet another anticompetitive, illegal measure brought to you by the experts on anticompetitive, illegal behavior: Applesoft. You can smell the desperation to keep the market from becoming performance-based, where they just can't compete.

• Michael

  2012-12-03 18:15:00

  Your interpretation of the law is flawed. It has never been deemed illegal to ship a full product with all components needed to use the tool even if you can replace parts of that tool. For example, I cannot buy the game "Monopoly" and then say I do not need the rules because I use my own open source rules and thus I should get a refund (or be able to buy the game sans rules). This does not mean a seller cannot sell the parts - they often do!
  
  In some industries there are special rules. For vehicles, for example, parts have to be available for X number of years. You can buy just a transmission or just the seats, for example. Maybe you can try to have those types of laws made to apply to computers. I doubt it though. Simply would not make as much sense.

• Dr. Roy Schestowitz

  2012-12-03 18:20:15

  Regarding USDOJ, I started to dislike it with a passion almost 2 years ago when it got involved in witch-hunting Wikileaks (potentially breaking laws in the process) and then again when it helped FBI illegally raid Megaupload, essentially helping a cartel (copyright cartel).
  
  Regarding UEFI, it is clearly anti-competitive and it is already preventing some people from moving to GNU/Linux. Now that Red Hat, LF, Canonical etc. missed their chance of filing antitrust complaint I guess we can rely on Shim.
  
  

  • Garrett releases first-stage bootloader to facilitate secure boot

    All machines pre-loaded with Windows 8 come with secure boot turned on; this means that other operating systems cannot boot on these machines unless they also support the process.

  
  
  

  • Secure Boot bootloader for Linux

    Linux developer Matthew Garrett has released a version of his Shim Secure Boot bootloader that allows any Linux distribution to be launched on Secure Boot systems without the need to disable UEFI Secure Boot. As Garrett's Shim binary has been signed by Microsoft, the Secure Boot bootloader will be executed by almost any type of UEFI firmware.

  
  
  Shim would harm attempts at antitrust complaint.

• Dr. Roy Schestowitz

  2012-12-03 21:59:14

  With Shim, Microsoft can also make it hard to boot Linux e.g. in case of war when it willingly (or by orders from the government) invalidates keys
  
  Also see Has Secure Boot for Linux Finally Arrived? from the 'Thank You Microsoft for *allowing* Linux on Windows 8 hardware?!' dept.

• mjg59

  2012-12-03 22:10:17

  If you don't trust Microsoft not to revoke binaries that you depend on, don't install the Microsoft revocation list updates. They can't blacklist binaries unless you choose to accept the update.

• Dr. Roy Schestowitz

  2012-12-03 22:14:34

  What will the default setting be? For a project that's independent from Microsoft (and actively attacked by Microsoft, e.g. using software patents, EDGI) it's unwise to be dependent on keys/certificates from Microsoft, whose track record is dubious at best.

• Michael

  2012-12-03 22:19:25

  By hardware that does not come with an MS OS.

• mjg59

  2012-12-03 22:30:02

  The default behaviour will be whatever your distribution decides it should be, but in all cases you'll be able to override it.

• mjg59

  2012-12-03 22:13:16

  "Shim would harm attempts at antitrust complaint."
  
  It's always been obvious that something like Shim was technically possible, and if Shim was unavailable it would be because people had chosen not to implement it. Are you asserting that whether or not something is illegally anti-competitive depends on whether or not other vendors choose to write something?

• Dr. Roy Schestowitz

  2012-12-03 22:20:45

  No, I am just saying that Microsoft is allowed to get away with damage it has already done, continues to do, and will do in the future using UEFI. It's a deterrent against BSD and Linux. Complaints were in order.
  
  "OpenBSD founder Theo de Raadt has slammed Red Hat and Canonical for the way they have reacted to Microsoft's introduction of "secure" boot along with Windows 8, describing both companies as wanting to be the new Microsoft."
  
  http://www.itwire.com/business-it-news/open-source/55924-openbsds-de-raadt-slams-red-hat-canonical-over-secure-boot

• mjg59

  2012-12-03 22:31:09

  So how does Shim harm attempts at complaints? If the complaints were valid before Shim, they're valid after it.

• Michael

  2012-12-03 22:35:37

  I do not see OS X users whining as you do, Roy.

• Dr. Roy Schestowitz

  2012-12-03 22:42:28

  Shim lets Microsoft get away with the claim that "Linux already deals with the issue" or something along those lines. It's a legitimiser, just like Novell's patent deal (2006) and other Microsoft incursions that let it pretend it's not FOSS opponent and patent aggressor.

• mjg59

  2012-12-03 23:09:08

  And if nobody had written it, Microsoft could just claim that someone should write it.

• Dr. Roy Schestowitz

  2012-12-03 23:25:32

  That's like arguing that a Microsoft patent lawsuit it made OK because Microsoft says developers can 'work around' the patent.

• mjg59

  2012-12-03 23:34:54

  We're talking about whether Microsoft are behaving anti-competitively. The behaviour of companies other than Microsoft cannot change whether Microsoft's behaviour is legal or not. If a technical solution could be implemented such that Microsoft's behaviour isn't anti-competitive, then it doesn't matter whether or not anyone actually implemented it - the behaviour isn't anti-competitive. If it's anti-competitive before Shim, it's still anti-competitive afterwards. If it wasn't before, it still isn't. The existence of Shim doesn't change whether or not Microsoft are breaking the law.

• Jose_X

  2012-12-03 23:55:03

  >> If a technical solution could be implemented
  
  But how do you show that such a solution could be implemented? A long complex mathematical solution that no one is ever going to write, a very convincing simple argument that may exist, or a proof of concept of some sort. These are your main options.
  
  And the quality/features of the "proof-of-concept" product and how hard it is to create it can show significant competitive disadvantages and distortion of the playing field.
  
  I agree though that if shim made a particular anti-competitive problem void, then there would be nothing to prosecute against the alleged abuser.

• Dr. Roy Schestowitz

  2012-12-03 23:56:20

  The case regarding "anti-competitive" (as per definition) practices will be determined based on evidence in theory and practice.

• mjg59

  2012-12-04 00:03:22

  And if called as an expert witness, I would be obliged to testify that the implementation of a solution along the lines of shim would be straightforward for anyone sufficiently skilled in the arts of operating system development. It's conceptually obvious, and its existence doesn't alter whether or not Microsoft are guilty.

• Dr. Roy Schestowitz

  2012-12-04 00:07:53

  Such a testimony would help Microsoft.

• mjg59

  2012-12-04 00:09:54

  Such a testimony would be true, and lying in court is generally considered to be a Bad Thing.

• Michael

  2012-12-04 00:15:09

  Please keep in mind that Roy is completely irrational when it comes to such topics. He believes MS has done something illegal because desktop Linux has not earned a place on many desktops. Roy would deny it, but in essence he believe open source cannot do well on the desktop without MS's help.

• Michael

  2012-12-04 00:17:01

  "Such a testimony would help Microsoft." - Roy
  
  That shows Roy's level of moral development: if it helps MS it is bad, if it hurts MS it is good. It is not about what is intrinsically right or wrong, good or bad... it is about "winning" at any cost (other than actually making desktop Linux earn its spot on the desktop).

• Dr. Roy Schestowitz

  2012-12-04 00:21:58

  mjg59, the point I was making is, creating a circumstance as such would be a Bad Thing™.

• mjg59

  2012-12-04 00:25:00

  What, starting a lawsuit? I thought you wanted one of them?

• Dr. Roy Schestowitz

  2012-12-04 00:31:15

  Let me explain. What I meant was, if a case was made for antitrust investigation (not a lawsuit), then querying someone like you, e.g. to say UEFI had Shim, would be favourable to Microsoft's position. You chose to play along with discriminatory (towards Linux and BSD for the most part) plot rather than antagonise it.

• mjg59

  2012-12-04 00:34:16

  If I hadn't written Shim, I'd still know that it was straightforward. Are you saying that I should refuse to give full and truthful testimony to an antitrust investigation?

• Dr. Roy Schestowitz

  2012-12-04 00:43:14

  mjg59,
  
  No, you re using a straw man again. What I meant was, supporting what you claim to be trivial to support makes ammo for them -- ammo with which to dismiss the problem of their anti-competitive tactics.
  
  Let me take you back to 2007. Microsoft had paid Novell to support OOXML and did similar things with other companies. At the same time a lot of complaints were made against OOXML, alleging -- correctly -- that it was a bogus 'open' format with patents and no working implementations, not even by Microsoft. By having Novell as an essentially "bribed supporter" Microsoft had the ammo with which to silence opposition while fighting against ODF and open standards policies in entire nations. I covered this in hundreds of posts in this site. What you do is akin to what GNOME/Gnumeric did at the time, not just Go-OO.

• mjg59

  2012-12-04 00:48:35

  You're missing the point. You said “Shim would harm attempts at antitrust complaint”. That's not true. The fact that it's possible for Shim to be written might harm attempts at antitrust complaint, the fact that someone actually wrote it wouldn't.

• Dr. Roy Schestowitz

  2012-12-04 01:00:00

  I think we had a case of miscommunication. I'll clarify further.
  
  Picture this hypothetical scenario. $LinuxVendor approaches USDOJ, alleging that X customers are unable to install GNU/Linux on Y new machines. $LinuxVendor shows that Microsoft produced requirements that led OEMs to tying an OS, Windows, to their hardware. Microsoft, when queried by USDOJ, is unable to justify those requirements. Torvalds alleges it does little for security, OpenBSD's founder openly complains about Red Hat, and Microsoft then cites a (former?) Red Hat developer who has said UEFI is all fine and dandy. This does not change the fact that Microsoft cannot justify those aforementioned requirements. This is similar not only to the OOXML situation but also to FAT and DRM. Microsoft is trying to 'normalise' anti-competitive tactics and those who play along become unhelpful to the victims. They help Microsoft ram down people's throats (and through regulators' door) the thing that reduces their control over their computing while also harming competition (see Freiburg's story).
  
  Microsoft has done this before. I covered it. I showed it. UEFI is more of the same.

• Michael

  2012-12-04 01:02:47

  Let me explain. What I meant was, if a case was made for antitrust investigation (not a lawsuit), then querying someone like you, e.g. to say UEFI had Shim, would be favourable to Microsoft’s position. You chose to play along with discriminatory (towards Linux and BSD for the most part) plot rather than antagonise it.

  
  
  Sad, Roy, how you think being honest and accurate to the best of his ability is a bad thing. Sickening, really. But at least you are making it clear how you decide what is moral: to you the ends justify the means - lying is fine to you. It has been clear for a long time but I think this is the first time I have seen you admit to it.

• mjg59

  2012-12-04 01:17:19

  No, let's go back to what you said. “Shim would harm attempts at antitrust complaint” - that's untrue. The concept behind Shim could harm attempts at antitrust complaints, because the mere fact that it's possible could make the difference between illegally anti-competitive behaviour and perfectly legal behaviour. The problem is, the concept is pretty obvious to anyone who spends some time thinking about the problem. Microsoft wouldn't have to rely on me. They could find many people who would testify that it would be possible for Linux vendors to implement a solution that satisfied their requirements. The only thing that's made an antitrust complaint more difficult is reality, not me.

• Dr. Roy Schestowitz

  2012-12-04 01:22:31

  Actively embracing discriminatory new 'technology' is the problem, especially when done by so-called "FOSS people" (I put "so-called" there because Microsoft uses bribed companies like Novell/SUSE and fakers like 'FOSSpatents' for this). It's hardly different from the OOXML scenario of 2007. I can elaborate if you want.

• mcinsand

  2012-12-03 20:50:18

  Regarding the DoJ, how can we expect the rest of the world to respect our laws if we don't respect them, whether we are concerned with detaining people for indefinite periods, trumping up charges to hijack other countries' citizens, or daily tolerating Applesoft's efforts to illegally tie hardware to software?
  
  On the topic of tying, there are some great analyses out there, and I got to work my way through a couple over lunch. Basically, the requirements for illegal tying are that a company have enough of a market presence to have some impact on the market dynamics, for component A to have some value, and then for component B to have value, too. It was meeting these three critiria that caused the SCOTUS to rule that DG tying RDOS to the Nova was illegal. DG didn't have a monopoly or even a majority market share, but they had enough visibility for their hardware and software to have value. Through this lens, MS' IE games are worse, but it is still illegal.
  
  Although we have historically been thinking of tying issues with respect to to desktops and laptops, tablets and cellphones are certainly more powerful than the computers in the market when IE was tied to Windows95, and particularly more powerful than the DG Nova (I know; I used a Nova 4 for much of the late '80's.). Does the OS have a value? Does the hardware have a value? Does the person peddling the combination of OS and hardware have a significant market presence? Then, and here is the big question if all three conditions are true: is the supplier making this legal by offering hardware and software separately, rather than bundled?

• Michael

  2012-12-03 21:24:56

  What is illegal about selling a complete, working system? Oh. Nothing. There is not a thing "illegal" about it. Look at my example with Monopoly - are you whining about how you have to pay for the instructions?
  
  As far as the value: pretty much software has *no* value without hardware... and hardware has essentially no value without software. You need both to have a complete, working system.

• Dr. Roy Schestowitz

  2012-12-03 23:58:17

  mcinsand, note that a lot of people forget that different rules apply to a monopoly. This has been a source/root of much spin.

• mcinsand

  2012-12-04 00:37:28

  Roy, yes, there are different rules for monopoly, but antitrust law applies to more than just monopolies. In the DG versus Digidyne case, DG clearly did not have a monopoly, but they had an operating system that the market valued. A problem with a lot of the FOSS community, among others, has been the misunderstanding that anticompetitive behavior is okay, as long as a company has some other company that it can call competition or holds a majority market share. That is one reason I put in some of the quotes from the DG versus Digidyne ruling an analysis. These laws aren't just designed to address existing monopolies but to also make some effort at punishing companies that compete in unethical ways.
  
  The analyses that I went through today are indeed very interesting, especially a Duke law article on illegal bundling and tying, and it helped to go back through and reread the DG versus Digidyne rulings. Hardware and software are not just separate markets by virtue of common sense, but the courts have ruled so, as well.

• Dr. Roy Schestowitz

  2012-12-04 00:47:31

  Yes, but I was addressing FUD and responding to misdirection from some Microsoft apologists.
  
  I don't like analogies that involve Google/Android because Google does not mandate that manufacturers lock the bootloader to earn some kind of "certification" or special treatment. Moreover, the desktop market has a chokehold which other markets don't have and general-purpose computers should ban tying.

• Michael

  2012-12-04 01:11:35

  Picture this hypothetical scenario. $LinuxVendor approaches USDOJ, alleging that X customers are unable to install GNU/Linux on Y new machines. $LinuxVendor shows that Microsoft produced requirements that led OEMs to tying an OS, Windows, to their hardware.

  This would only apply to hardware with MS software. MS cannot make requirements for what is needed for Ubuntu!

  Microsoft, when queried by USDOJ, is unable to justify those requirements.

  Does Canonical have to justify their requirements?

  Torvalds alleges it does little for security, OpenBSD's founder openly complains about Red Hat, and Microsoft then cites a (former?) Red Hat developer who has said UEFI is all fine and dandy.

  MS does things to benefit its OS - things which other OS developers are not doing. The other OS developers say it does not do much. Do you see the potential conflict of interest? Do not get me wrong, I am not saying it is a good thing but, assuming you even are correct about your claims (you often are not), the claims are still weak evidence.

  This does not change the fact that Microsoft cannot justify those aforementioned requirements.

  Justify to your satisfaction, you mean. MS has plent of info on UEFI on their site.

  This is similar not only to the OOXML situation but also to FAT and DRM. Microsoft is trying to 'normalise' anti-competitive tactics and those who play along become unhelpful to the victims.

  You have a very strong victim mentality. You think the world owes you and the open source community things. It does not. You blame the boogieman for desktop Linux doing poorly. All very irrational. Right or wrong on this topic your support is very weak.

  They help Microsoft ram down people's throats (and through regulators' door) the thing that reduces their control over their computing while also harming competition (see Freiburg's story).

  Being a strong competitor and making products people want does "harm" competitors. That is an affect of choice.

  Microsoft has done this before. I covered it. I showed it. UEFI is more of the same.

  You make things up a lot and obsess over MS and Apple - pretending they are the bad guys to the good guys of Samsung and Google (for example).

• mcinsand

  2012-12-04 11:14:18

  Roy, I apologize if I was pushing the discussion in a direction that you didn't want to follow, but the topic of UEFI has me concerned, and I was excited yesterday after reading some of the rulings and analyses. Also, after the ordeal of Microsoft illegally comingling code between Windows95 and IE, well, UEFI is very much a repeat of the same pattern, in my book. Courts ruled that welding IE to Windows was illegal tying, thus ruling that applications and OS are separate markets, just as they ruled a decade earlier that OS and hardware are separate. Previously, the justification was for the user experience, and, now, MS is trying to tie software to hardware as a means to address the vulnerability industry created by MS' technical incompetence. History is repeating, and we are seeing the same anticompetitive measures all over again.
  
  I'm not sure where the Google/Android came in, but Google does look to be doing some things right, yet again. The EFI on Chromebooks first had the hairs on the back of my neck standing, but Google actually publicizes how to get full access needed for installing other OS's.

• mcinsand

  2012-12-04 11:16:15

  Roy, one more thing. I am pretty sure I know how to get this to you, but there is one particular piece on tying that I would really like to forward to you, if you're interested.

• Dr. Roy Schestowitz

  2012-12-04 11:19:15

  Am interested.

• Dr. Roy Schestowitz

  2012-12-04 11:48:02

  Google is not in the business of imposition, neither in search nor in OSes. What i don't like is, people send links to Docs/YouTube, whereupon (in the case of work) it becomes less of a choice. But it's usually avoidable.

• Michael

  2012-12-04 13:47:22

  You can opt to not use MS or Google, sure. Or Apple. Or Samsung. This is called choice.

• mcinsand

  2012-12-04 14:20:42

  Roy,
  
  Rather than try to find the information that I had before, here is a link to the article on computers and illegal tying. As a rare exercise of discipline, I will not go overlength here, although drafts of this message were considerably longer than just a few lines ;) This article has many points worth a paragraph on relevance to today's technical market.
  
  Regards.